import requests
import lyricsgenius as genius

nom_fichier = "genius_id.txt"
dico_token = {}

with open(nom_fichier, "r", encoding="utf-8") as fichier:
    for ligne in fichier:
        cle, valeur = ligne.split(":", 1)
        dico_token[cle.strip()] = valeur.strip()

#Connection à l'api et création de ma requête.
api=genius.Genius(dico_token["client_token_1"])
artist=api.search_artist('Rihanna', max_songs=2)

#Je ne veux que les métadonnées titres et paroles des chansons.
titles=[song.title for song in artist.songs]
lyrics=[song.lyrics for song in artist.songs]
thingstosave=[]
for i in range(0,len(artist)):
    thingstosave.append((titles[i],lyrics[i]))

thingstosave[0]

('Work',
 '[Chorus: Rihanna]\nWork, work, work, work, work, work\nHe said me haffi work, work, work, work, work, work\nHe see me do mi dirt, dirt, dirt, dirt, dirt, dirt\nAh so me put in work, work, work, work, work, work\nWhen you ah guh learn, learn, learn, learn, learn, learn?\nMe nuh cyar if him hurt, hurt, hurt, hurt, hurting\n\n[Verse 1: Rihanna]\nDry, me ah desert him\nNuh time to have you lurking\nHim ah go act like he nuh like it\nYou know I dealt with you the nicest\nNuh body touch me, you nuh righteous\nNuh badda, text me in a crisis\nI believed all of your dreams, adoration\nYou took my heart and my keys and my patience\nYou took my heart on my sleeve for decoration\nYou mistaken my love I brought for you for foundation\nAll that I wanted from you was to give me\nSomething that I never had\nSomething that you\'ve never seen\nSomething that you\'ve never been, mm\nBut I wake up and act like nothing\'s wrong\n\n[Chorus: Rihanna]\nJust get ready fi work, work, work, work, work, work\nHe said me haffi work, work, work, work, work, work\nHe see me do mi dirt, dirt, dirt, dirt, dirt, dirt\nAh so me put in work, work, work, work, work, work\nNer, ner, ner, ner, ner, ner\nWhen you ah guh learn, learn, learn, learn, learn, learn?\nBefore the tables turn, turn, turn, turn, turn, turn\n\n[Verse 2: Rihanna]\nBeg you something, please\nBaby, don\'t you leave\nDon\'t leave me stuck here in the streets, uh-huh\nIf I get another chance to\nI will never, no, never neglect you\nI mean, who am I to hold your past against you?\nI just hope that it gets to you\nI hope that you see this through\nI hope that you see this true\nWhat can I say?\nPlease recognize I\'m tryin\', babe\n\n[Chorus: Rihanna & Drake]\nI haffi work, work, work, work, work, work\nHe said me haffi work, work, work, work, work, work\nHe see me do mi dirt, dirt, dirt, dirt, dirt, dirt\nSo me put in work, work, work, work, work, work\nWhen you ah guh learn, learn, learn, learn, learn, learn?\nMe nuh cyar if him hurt, hurt, hurt, hurt, hurting (\nYeah, okay\n)\n\n[Verse 3: Drake, Rihanna, Both]\nYou need to get done, done, done, done at work, come over\nWe just need to slow the motion\nDon\'t give that away to no one\nLong distance, I need you (Ah)\nWhen I see potential, I just gotta see it through\nIf you had a twin, I would still choose you\nI don\'t wanna rush into it if it\'s too soon\nBut I \nknow you need to get done, done, done, done\nIf you come over\nSorry if I\'m way less friendly\nI got niggas tryna end me, oh (Yeah)\nI spilled all my emotions tonight, I\'m sorry\nRollin\', rollin\', rollin\', rollin\', rollin\'\nHow many more shots until you\'re rollin\'?\nWe just need a face-to-face\nYou could pick the time and the place\nYou spent some time away\nNow you need to forward\n\n[Chorus: Rihanna, Rihanna & Drake]\nAnd give me all the work, work, work, work, work\n, work\nHe said me haffi \nwork, work, work, work, work, work\nHe see me do mi dirt, dirt, dirt, dirt, dirt, dirt\nSo me put in work, work, work, work, work, work\nWhen you ah guh learn, learn, learn, learn, learn, learn?\nMe nuh cyar if him hurt, hurt, hurt, hurt, hurting\n\n[Outro: Rihanna, Rihanna & Drake]\nMm\nMm\nWork, work, work, work, work, work\nMm\n\n[Click here to learn more about the making of "Work"]\n\n[Produced by Boi-1da]')

liste_token = ["client_token_"+ str(i) for i in range(1,4)]

liste_artiste = ["Bruno Mars","Rihanna","Beyonce","Alicia Keys","Frank Ocean",
                 "Amy Winehouse","Justin Timberlake", "Usher", "Mariah Carey", "Justin Bieber","Michael Jackson",
                 "Christina Aguilera", "Chris Brown", "Ne-Yo", "Destiny's Child", "Aaliyah", "Mary J. Blige",
                 "Marvin Gaye", "Janet Jackson", "Stevie Wonder"]

dico_artiste_token = {"client_token_1" : liste_artiste[:7],
                      "client_token_2" : liste_artiste[7:14],
                      "client_token_3" : liste_artiste[14:]}
dico_artiste_token

{'client_token_1': ['Bruno Mars',
  'Rihanna',
  'Beyonce',
  'Alicia Keys',
  'Frank Ocean',
  'Amy Winehouse',
  'Justin Timberlake'],
 'client_token_2': ['Usher',
  'Mariah Carey',
  'Justin Bieber',
  'Michael Jackson',
  'Christina Aguilera',
  'Chris Brown',
  'Ne-Yo'],
 'client_token_3': ["Destiny's Child",
  'Aaliyah',
  'Mary J. Blige',
  'Marvin Gaye',
  'Janet Jackson',
  'Stevie Wonder']}

# Fonction permettant de récupérer n_chansons d'un artiste et de les lister dans un dictionnaire
def telecharger_paroles(api,artiste,n_chansons = 200,download_each_file = False):
    dico_artiste = {}
    artiste_info=api.search_artist(artiste, max_songs=n_chansons)
    dico_artiste = {song.title : song.lyrics for song in artiste_info.songs}
    if download_each_file :
        nom_fichier = artiste.replace(" ","_").replace(",","").replace("&","").replace("-"," ") + ".txt"
        with open(nom_fichier, 'w') as outfile:
            json.dump(dico_artiste, outfile)
    return dico_artiste

# Fonction qui va être paralléliser et permettant de se connecter à l'api Genius afin de créer un dictionnaire par artiste.
data_json = {}
def connect_api(Token,artistes, n_chansons = 200):
    str_token = dico_token[Token]
    api=genius.Genius(str_token,timeout=20, retries=5)
    for artiste in artistes[Token]:
        dico_artiste = telecharger_paroles(api,artiste, n_chansons = n_chansons)
        data_json[artiste] = dico_artiste

from concurrent.futures import ThreadPoolExecutor
import json

with ThreadPoolExecutor(3) as pool:
  values = pool.map(lambda x : connect_api(x,dico_artiste_token),liste_token)

with open("data_json.json", 'w') as outfile:
    json.dump(data_json, outfile)

import collections
import re
from IPython.display import HTML, display

def supp_doublons_chansons(dico, artiste):
    titres_bruts = list(dico[artiste].keys())
    
    # On créé la Regex pour identifier la partie "variante" à nettoyer
    pattern_nettoyage = re.compile(
        r"\s*(?:[-([][^([)]*(?:remix|edit|live|acoustic|version|radio|mix|ft\.|feat\.)[^([)]*[-)\]]?)",
        flags=re.IGNORECASE,
    )
    
    # On regroupe tous les titres originaux par leur nom "propre"
    groupes_titres = collections.defaultdict(list)
    for titre_original in titres_bruts:
        titre_propre = pattern_nettoyage.sub("", titre_original).strip().lower()
        groupes_titres[titre_propre].append(titre_original)
    
    # Pour chaque groupe, on choisit le MEILLEUR exemplaire original
    # Comme on veut la version épurée, on part du principe que ca sera toujours la version la plus courte
    titres_finaux_conserves = []
    
    for titre_propre, liste_originaux in groupes_titres.items():
        titre_court = min(liste_originaux, key=len, default="")
        titres_finaux_conserves.append(titre_court)
        
    # On affiche les titres écartés
    titres_ecartes = sorted(set(titres_bruts).difference(set(titres_finaux_conserves)))

    if titres_ecartes:
        contenu_html = f"<p>Les chansons de l'artiste <b>{artiste}</b> qui ont été supprimées sont les suivantes :</p>"
        contenu_html += """<details>
        <summary>Afficher la liste des titres supprimés</summary>
        <ul>"""
        for element in titres_ecartes:
            contenu_html += f"<li>{element}</li>"
        contenu_html += "</ul></details>"
    else:
        contenu_html = f"<p>Aucune chanson écartée pour l'artiste {artiste}.</p>"
    display(HTML(contenu_html))

    dico_artiste_sans_doublons = {chanson:paroles for chanson, paroles in dico[artiste].items() if chanson in titres_finaux_conserves}
    return dico_artiste_sans_doublons

corpus = {}
for artiste in data_json:
    dico_artiste_sans_doublons = supp_doublons_chansons(data_json,artiste)
    corpus[artiste] = dico_artiste_sans_doublons

with open("corpus.json", 'w') as outfile:
    json.dump(corpus, outfile)

import json
from unidecode import unidecode

import nltk
from nltk.stem import PorterStemmer
from nltk.stem import WordNetLemmatizer

wnl = WordNetLemmatizer()
stm = PorterStemmer()
with open("corpus.json",encoding="utf-8") as f:
    corpus = json.load(f)
    
import re
import string

corpus ={cle:val for cle, val in corpus.items()}

# Extraction et aplatissement des Titres (title_song_tool), des Données (X) et des Artistes (Y)
X = []
titres_bruts = []
Y = []

for artiste, chansons in corpus.items():
    for titre, paroles in chansons.items():
        titres_bruts.append(titre)
        X.append(paroles)
        Y.append(artiste)

title_song_tool = {i: titre for i, titre in enumerate(titres_bruts)}

print(len(title_song_tool),"chansons importées")

3193 chansons importées

import re
pattern_variante = re.compile(
    r"\s*(?:[-([][^([)]*(?:remix|edit|live|acoustic|version|radio|mix|ft\.|feat\.)[^([)]*[-)\]]?)",
    flags=re.IGNORECASE)

titres_variants_trouves = [titre for titre in title_song_tool.values() if re.search(pattern_variante, titre)]
compteur_variantes = len(titres_variants_trouves)

# On affiche les résultats
print(f"Nombre de chansons variantes : {compteur_variantes}")
contenu_html = """<details>
<summary>Afficher la liste des chansons variantes détectées</summary>
<ul>"""
for titre in titres_variants_trouves:
    contenu_html += f"<li>{titre}</li>"
contenu_html +="</ul></details>"
display(HTML(contenu_html))

Nombre de chansons variantes : 87

def cleaner(chanson):
    # On passe en minuscules, on vire les accents et on remplace l'apostrophe courbe ’ de Genius par une apostrophe droite '
    chanson = unidecode(chanson).lower()
    chanson = chanson.replace("’", "'")

    # Nettoyage des bruits spécifiques de Genius (Balises et boutons de partage)
    chanson = re.sub(r"\[.*?\]", " ", chanson)
    chanson = re.sub(r"\d*\s*embed\s*share.*", " ", chanson)
    chanson = re.sub(r"embed\s*copy", " ", chanson)

    # Normalisation des contractions anglaises courantes
    chanson = re.sub(r"n't\b", " not", chanson)
    chanson = re.sub(r"\bi'm\b", "i am", chanson)
    chanson = re.sub(r"'s\b", " is", chanson)
    chanson = re.sub(r"'ll\b", " will", chanson)
    chanson = re.sub(r"'ve\b", " have", chanson)
    chanson = re.sub(r"'re\b", " are", chanson)
    chanson = re.sub(r"'d\b", " would", chanson)

    # On supprime la ponctuation
    ponctuations = string.punctuation + "\n\r"
    chanson = re.sub(f"[{re.escape(ponctuations)}]", " ", chanson)

    # Nettoyage final des espaces multiples
    chanson = re.sub(r"\s+", " ", chanson)

    return chanson.strip()

values = map(cleaner,X)
values = list(values)

who_do_we_remove = {idx:(Y[idx],title_song_tool[idx]) for idx,valeur in enumerate(values) if len(valeur) <= 20}
print(f"{len(who_do_we_remove)} chansons ont été supprimées car il y a moins de 20 mots dans leur texte.")

30 chansons ont été supprimées car il y a moins de 20 mots dans leur texte.

print("la liste des chansons qui ont été supprimées est la suivante :")
for i in who_do_we_remove.keys():
    print(i,":",title_song_tool[i])

la liste des chansons qui ont été supprimées est la suivante :
130 : So Emotional
138 : S.Y.T (Sweet, Young & Tender)
141 : Intro/Overture
142 : Soldier (Dance Interlude)
145 : Life
407 : Don’t Call Me No More
429 : Back & Forth (Ms. Mello Instrumental)
1418 : Main Theme From Trouble Man (2)
1434 : “T” Plays It Cool
1441 : I Want You (Intro Jam)
1444 : T’' Stands For Trouble
1940 : Interlude: Online
1959 : Are You Still Up
1967 : Sweet Dreams
1978 : Back
1981 : Interlude: Come Back
2158 : Kimono Girl
2321 : As I Am (Intro)
2325 : De Novo Adagio
2348 : KEYS
2619 : In Here Somewhere
2625 : White
2626 : Rushes To
2636 : Florida
2995 : Over You
3016 : New Love
3152 : Fresh Leaves
3177 : People Everyday
3185 : She Likes The Rain
3188 : As Dreamers Do

Y  = [artist for i,artist in  enumerate(Y) if i not in who_do_we_remove.keys()]
X  = [lyrics for i,lyrics in  enumerate(values) if i not in who_do_we_remove.keys()]

title_song_tool = {cle:valeur for cle, valeur in title_song_tool.items() if cle not in list(who_do_we_remove.keys())}
## On vérifie que ca a bien supprimé
print(len(title_song_tool))

3163

from nltk.stem import PorterStemmer
from nltk.corpus import stopwords
english_stopwords = set(stopwords.words("english"))

def stem_stopwords(chanson,liste_motvides):
    return(
        [stm.stem(mot) for mot in chanson.split() if mot not in liste_motvides]
    )

X = [stem_stopwords(chanson,english_stopwords) for chanson in X]

from sklearn.feature_extraction.text import TfidfVectorizer

def dummy_fun(doc):
    return doc

vectorizer = TfidfVectorizer(ngram_range=(1, 2),
                             use_idf=True, smooth_idf=True, # idf lissé
                             min_df=2,
                             sublinear_tf=True,
                             norm='l2',
                             analyzer = 'word',
                             tokenizer=dummy_fun,
                             preprocessor = dummy_fun)

dtm = vectorizer.fit_transform(X)

C:\Users\User\anaconda3\envs\projet_musique\Lib\site-packages\sklearn\feature_extraction\text.py:526: UserWarning: The parameter 'token_pattern' will not be used since 'tokenizer' is not None'
  warnings.warn(

import pandas as pd
pd.DataFrame(dtm.toarray(),columns=vectorizer.get_feature_names_out())

from sklearn.model_selection import train_test_split,GridSearchCV
from sklearn.naive_bayes import MultinomialNB

x_train,x_test,y_train,y_test = train_test_split(dtm,Y,test_size = 0.33)

MultinomialNB().get_params()

{'alpha': 1.0, 'class_prior': None, 'fit_prior': True, 'force_alpha': True}

import numpy as np

mod= MultinomialNB()

parametres = {
    "alpha": [0.001,0.005,0.01,0.05,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0],
    "fit_prior" : [True, False],

}
grid = GridSearchCV(mod,parametres,verbose=1,scoring = "roc_auc_ovr_weighted",n_jobs=4)
grid.fit(x_train,y_train)

Fitting 5 folds for each of 28 candidates, totalling 140 fits

GridSearchCV(estimator=MultinomialNB(), n_jobs=4,
             param_grid={'alpha': [0.001, 0.005, 0.01, 0.05, 0.1, 0.2, 0.3, 0.4,
                                   0.5, 0.6, 0.7, 0.8, 0.9, 1.0],
                         'fit_prior': [True, False]},
             scoring='roc_auc_ovr_weighted', verbose=1)

MultinomialNB(alpha=0.2, fit_prior=False)

import matplotlib.pyplot as plt

plt.scatter(y= grid.cv_results_["mean_test_score"],
            x = grid.cv_results_["param_alpha"].data,
            color="blue",
            edgecolor="k",
            alpha=0.7)
plt.title(
    "Impact du paramètre de régularisation (Alpha) sur le score de validation",
    fontsize=12,
    fontweight="bold",
    pad=15,
)
plt.xlabel("Valeur à l'échelle logarithmique de l'Hyperparamètre Alpha", fontsize=10)
plt.ylabel("Score ROC AUC moyen (OVR Pondéré)", fontsize=10)

plt.xscale("log")

plt.grid(True, which="both", linestyle="--", alpha=0.5)
plt.show()

grid.best_params_

{'alpha': 0.2, 'fit_prior': False}

prediction = grid.predict(x_test)

from sklearn.metrics import classification_report, roc_auc_score
roc_auc_weighted = roc_auc_score(
    y_test,
    grid.predict_proba(x_test),
    multi_class="ovr",
    average="weighted",
)

print(classification_report(y_test, prediction))
print("Nombre d'erreur de prédictions sur un total de %d observations : %d, ce qui représente %.3f de précision." % (x_test.shape[0], (y_test != prediction).sum(),(1-(y_test != prediction).sum()/x_test.shape[0])))
print(f"Le Score ROC AUC moyen (OVR Pondéré) est de : {roc_auc_weighted:.3f}")

                    precision    recall  f1-score   support

           Aaliyah       1.00      0.38      0.55        37
       Alicia Keys       0.81      0.21      0.34        61
     Amy Winehouse       0.94      0.33      0.48        49
           Beyonce       0.50      0.09      0.16        64
        Bruno Mars       1.00      0.11      0.20        37
       Chris Brown       0.17      0.88      0.28        56
Christina Aguilera       0.58      0.26      0.36        68
   Destiny's Child       0.77      0.17      0.28        58
       Frank Ocean       0.62      0.30      0.41        66
     Janet Jackson       0.52      0.26      0.35        53
     Justin Bieber       0.21      0.27      0.23        60
 Justin Timberlake       0.56      0.10      0.18        48
      Mariah Carey       0.24      0.38      0.30        55
       Marvin Gaye       0.41      0.92      0.57        53
     Mary J. Blige       0.24      0.50      0.33        58
   Michael Jackson       0.63      0.54      0.58        48
             Ne-Yo       0.45      0.33      0.38        64
           Rihanna       0.33      0.09      0.14        46
             Usher       0.26      0.30      0.28        63

          accuracy                           0.34      1044
         macro avg       0.54      0.34      0.34      1044
      weighted avg       0.52      0.34      0.33      1044

Nombre d'erreur de prédictions sur un total de 1044 observations : 690, ce qui représente 0.339 de précision.
Le Score ROC AUC moyen (OVR Pondéré) est de : 0.850

lyrics = """Hello, it's me I was wondering if after all these years you'd like to meet To go over everything 
They say that time's supposed to heal ya But I ain't done much healing 
Hello, can you hear me? I'm in California dreaming about who we used to be
When we were younger and free 
I've forgotten how it felt before the world fell at our feet
There's such a difference between us
And a million miles
Hello from the other side
I must've called a thousand times
To tell you I'm sorry for everything that I've done
But when I call, you never seem to be home
Hello from the outside
At least I can say that I've tried
To tell you I'm sorry for breaking your heart
But it don't matter, it clearly doesn't tear you apart anymore
Hello, how are you?
It's so typical of me to talk about myself, I'm sorry
I hope that you're well
Did you ever make it out of that town where nothing ever happened?
It's no secret that the both of us are running out of time
So hello from the other side (Other side)
I must've called a thousand times (Thousand times)
To tell you I'm sorry for everything that I've done
But when I call, you never seem to be home
Hello from the outside (Outside)
At least I can say that I've tried (I've tried)
To tell you I'm sorry for breaking your heart
But it don't matter, it clearly doesn't tear you apart anymore
Ooh-ooh, anymore
Ooh-ooh-ooh, anymore
Ooh-ooh, anymore
Anymore
Hello from the other side (Other side)
I must've called a thousand times (Thousand times)
To tell you I'm sorry for everything that I've done
But when I call, you never seem to be home
Hello from the outside (Outside)
At least I can say that I've tried (I've tried)
To tell you I'm sorry for breaking your heart
But it don't matter, it clearly doesn't tear you apart anymore"""

lyrics = cleaner(lyrics)
lyrics = stem_stopwords(lyrics,english_stopwords)

lyrics_dtm = vectorizer.transform(lyrics)
who_does_adele_sound_like = grid.predict_proba(lyrics_dtm)
pd.DataFrame({"artiste":grid.classes_,"probabilité":who_does_adele_sound_like[0]}).sort_values(by = "probabilité",ascending = False).head()

lyrics = """
I see trees of green, red roses too
I see them bloom for me and you
And I think to myself what a wonderful world.
I see skies of blue and clouds of white
The bright blessed day, the dark sacred night
And I think to myself what a wonderful world.
The colors of the rainbow so pretty in the sky
Are also on the faces of people going by
I see friends shaking hands saying how do you do
They're really saying I love you.
I hear babies crying, I watch them grow
They'll learn much more than I'll never know
And I think to myself what a wonderful world
Yes I think to myself what a wonderful world.
"""

lyrics = cleaner(lyrics)
lyrics = stem_stopwords(lyrics,english_stopwords)

lyrics_dtm = vectorizer.transform(lyrics)
who_does_adele_sound_like = grid.predict_proba(lyrics_dtm)
pd.DataFrame({"artiste":grid.classes_,"probabilité":who_does_adele_sound_like[0]}).sort_values(by = "probabilité",ascending = False).head(5)

	estimator estimator: estimator object This is assumed to implement the scikit-learn estimator interface. Either estimator needs to provide a ``score`` function, or ``scoring`` must be passed.	MultinomialNB()
	param_grid param_grid: dict or list of dictionaries Dictionary with parameters names (`str`) as keys and lists of parameter settings to try as values, or a list of such dictionaries, in which case the grids spanned by each dictionary in the list are explored. This enables searching over any sequence of parameter settings.	{'alpha': [0.001, 0.005, ...], 'fit_prior': [True, False]}
	scoring scoring: str, callable, list, tuple or dict, default=None Strategy to evaluate the performance of the cross-validated model on the test set. If `scoring` represents a single score, one can use: - a single string (see :ref:`scoring_string_names`); - a callable (see :ref:`scoring_callable`) that returns a single value; - `None`, the `estimator`'s :ref:`default evaluation criterion ` is used. If `scoring` represents multiple scores, one can use: - a list or tuple of unique strings; - a callable returning a dictionary where the keys are the metric names and the values are the metric scores; - a dictionary with metric names as keys and callables as values. See :ref:`multimetric_grid_search` for an example.	'roc_auc_ovr_weighted'
	n_jobs n_jobs: int, default=None Number of jobs to run in parallel. ``None`` means 1 unless in a :obj:`joblib.parallel_backend` context. ``-1`` means using all processors. See :term:`Glossary ` for more details. .. versionchanged:: v0.20 `n_jobs` default changed from 1 to None	4
	refit refit: bool, str, or callable, default=True Refit an estimator using the best found parameters on the whole dataset. For multiple metric evaluation, this needs to be a `str` denoting the scorer that would be used to find the best parameters for refitting the estimator at the end. Where there are considerations other than maximum score in choosing a best estimator, ``refit`` can be set to a function which returns the selected ``best_index_`` given ``cv_results_``. In that case, the ``best_estimator_`` and ``best_params_`` will be set according to the returned ``best_index_`` while the ``best_score_`` attribute will not be available. The refitted estimator is made available at the ``best_estimator_`` attribute and permits using ``predict`` directly on this ``GridSearchCV`` instance. Also for multiple metric evaluation, the attributes ``best_index_``, ``best_score_`` and ``best_params_`` will only be available if ``refit`` is set and all of them will be determined w.r.t this specific scorer. See ``scoring`` parameter to know more about multiple metric evaluation. See :ref:`sphx_glr_auto_examples_model_selection_plot_grid_search_digits.py` to see how to design a custom selection strategy using a callable via `refit`. See :ref:`this example ` for an example of how to use ``refit=callable`` to balance model complexity and cross-validated score. .. versionchanged:: 0.20 Support for callable added.	True
	cv cv: int, cross-validation generator or an iterable, default=None Determines the cross-validation splitting strategy. Possible inputs for cv are: - None, to use the default 5-fold cross validation, - integer, to specify the number of folds in a `(Stratified)KFold`, - :term:`CV splitter`, - An iterable yielding (train, test) splits as arrays of indices. For integer/None inputs, if the estimator is a classifier and ``y`` is either binary or multiclass, :class:`StratifiedKFold` is used. In all other cases, :class:`KFold` is used. These splitters are instantiated with `shuffle=False` so the splits will be the same across calls. Refer :ref:`User Guide ` for the various cross-validation strategies that can be used here. .. versionchanged:: 0.22 ``cv`` default value if None changed from 3-fold to 5-fold.	None
	verbose verbose: int Controls the verbosity: the higher, the more messages. - >1 : the computation time for each fold and parameter candidate is displayed; - >2 : the score is also displayed; - >3 : the fold and candidate parameter indexes are also displayed together with the starting time of the computation.	1
	pre_dispatch pre_dispatch: int, or str, default='2n_jobs' Controls the number of jobs that get dispatched during parallel execution. Reducing this number can be useful to avoid an explosion of memory consumption when more jobs get dispatched than CPUs can process. This parameter can be: - None, in which case all the jobs are immediately created and spawned. Use this for lightweight and fast-running jobs, to avoid delays due to on-demand spawning of the jobs - An int, giving the exact number of total jobs that are spawned - A str, giving an expression as a function of n_jobs, as in '2n_jobs'	'2*n_jobs'
	error_score error_score: 'raise' or numeric, default=np.nan Value to assign to the score if an error occurs in estimator fitting. If set to 'raise', the error is raised. If a numeric value is given, FitFailedWarning is raised. This parameter does not affect the refit step, which will always raise the error.	nan
	return_train_score return_train_score: bool, default=False If ``False``, the ``cv_results_`` attribute will not include training scores. Computing training scores is used to get insights on how different parameter settings impact the overfitting/underfitting trade-off. However computing the scores on the training set can be computationally expensive and is not strictly required to select the parameters that yield the best generalization performance. .. versionadded:: 0.19 .. versionchanged:: 0.21 Default value was changed from ``True`` to ``False``	False

Pousser la chansonnette : Classification et stylométrie de paroles de pop¶

Introduction ¶

Table des matières¶

Création du corpus ¶

Utilisation du Package lyricsgenius ¶

Importation du corpus ¶

Suppression des doublons ¶

Enregistrement du Corpus ¶

Preprocessing ¶

Restructuration de la base ¶

Pipeline de traitement automatique de la base ¶

Vectorisation ¶

Modélisation ¶

Choix du modèle ¶

Optimisation des Hyperparamètres ¶

Evaluation du modèle ¶

Prédiction ¶

"Hello" de Adele ¶

"What a Wonderful World" de Louis Armstrong ¶

Conclusion ¶

Documentation ¶

	0	000	01	01 know	1	1 1	1 2	10	10 9	10 minut	...	zone caus	zone got	zone move	zoom	zoom zoom	zoomin	zz	zz bee	zz come	zz zz
0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
1	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
2	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
3	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
4	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
3158	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
3159	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
3160	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
3161	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
3162	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0

	alpha alpha: float or array-like of shape (n_features,), default=1.0 Additive (Laplace/Lidstone) smoothing parameter (set alpha=0 and force_alpha=True, for no smoothing).	0.2
	force_alpha force_alpha: bool, default=True If False and alpha is less than 1e-10, it will set alpha to 1e-10. If True, alpha will remain unchanged. This may cause numerical errors if alpha is too close to 0. .. versionadded:: 1.2 .. versionchanged:: 1.4 The default value of `force_alpha` changed to `True`.	True
	fit_prior fit_prior: bool, default=True Whether to learn class prior probabilities or not. If false, a uniform prior will be used.	False
	class_prior class_prior: array-like of shape (n_classes,), default=None Prior probabilities of the classes. If specified, the priors are not adjusted according to the data.	None

	artiste	probabilité
3	Beyonce	0.118202
18	Usher	0.100323
11	Justin Timberlake	0.082943
0	Aaliyah	0.082004
5	Chris Brown	0.075223

	artiste	probabilité
10	Justin Bieber	0.090341
15	Michael Jackson	0.080668
3	Beyonce	0.076520
4	Bruno Mars	0.064778
5	Chris Brown	0.062814